The objective of this research is to investigate methods for providing reliable transmission of compressed information - speech, images, video, and data - over the bandlimited, fading channels that characterize mobile communication systems. The research will focus on two (intersecting) lines of investigation: (i) Error control strategies that exploit the properties of data compression algorithms: The goal here will be to make effective use of the "residual redundancy" remaining in signals after compression; virtually all presently-used source coding schemes leave some such redundancy, and channel decoders that are "tuned" to exploit it can offer substantial gain over decoders that are not. (Recent results regarding CELP speech coding show 2-5 dB of gain.) (ii)The design analysis, and implementation of error control codes specifically for digital mobile communication. It is well known that the distance properties that make a code "good" for additive white Gaussian noise channels are not the most important design properties for codes to be used over fading channels. Recently, the P.I. and collaborators have developed an approach to code design for fading channels that results in significantly improved performance with respect to the "traditional" approach; this research will extend the work in this area, developing new codes and analyzing their performance in a variety of contexts - including the synergy between the codes and "diversity reception" (i.e., multiple antennas) and the performance of the codes at low signal-to-noise ratios. The work to be undertaken will be a mixture of code design, analysis, and simulation. Successful completion of the proposed research will result in a capability to support multi-media transport over mobile communication channels.
